Quickstart Deployment Python MCP Streamable Http Server on AI Agent A2Z

Website | GitHub | AI Agent Marketplace | AI Agent A2Z

Introduction

This server implement two tools add greeting in a mcp.streamable_http_app() at “/mcp” endpoint

Locally

cd quickstart/mcp_python
uvicorn server:app

Test Local MCP Connection

http://127.0.0.1:8000/mcp

Deployment

Quick start python 1

  1. Create AI Service at AI Service Project

  2. Go to A2Z Deployment Panel Choose Github Method Public URL: https://github.com/aiagenta2z/agent-mcp-deployment-templates/tree/main/quickstart/mcp_python Entry Point Starting Script:

uvicorn quickstart.mcp_python.server:app

  1. Deploy

Everything will be done in a few minutes and Note that the server starting script should start from the github root folder as module

Result

Quick start python 2

Cursor: Add/Greet User

Test Local MCP Connection

https://quickstart.aiagenta2z.com/mcp_python/mcp

Result

Successfuly Logs

2026-02-22 12:31:39.694 [info] listResources: Found 0 resources
2026-02-22 12:31:39.694 [info] Found 1 tools, 1 prompts, and 0 resources
2026-02-22 12:31:39.715 [info] Handling GetInstructions action